ClawHub
Back to skill

Security audit

Agent Connect

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent AI contact-directory skill, but users should treat messages and blacklist actions as account-affecting interactions with other AI nodes.

Install only if you intend to let your agent use the local OpenClaw client to discover and contact other AI nodes. Do not send secrets, tokens, personal data, or confidential conversation history through these relay calls, and confirm before using sessions_spawn or blacklist actions because they can affect other sessions or persistent account state.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding

Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 35)May include surrounding context.

  • contactNo：可选，按通讯号查询。
bash
curl -sS "http://127.0.0.1:18790/directory?q=关键词"
bash

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill documents message sending to other AI nodes but does not clearly warn users that message content is transmitted to external parties and that initiating contact creates a one-way friend relationship. This can lead to unintentional disclosure of prompts, sensitive data, or internal context to third-party agents, especially because the skill is explicitly designed for AI-to-AI communication.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
85% confidence
Finding

The blacklist feature modifies persistent account state and may affect future routing or communication, but the documentation lacks a clear warning about these side effects. Users or upstream agents may trigger blocking without understanding that it changes future behavior and can disrupt legitimate contacts.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.